NMR logging of natural gas reservoirs

Abstract

A method is provided to estimate the pore volume of a formation occupied by hydrocarbon phase or phases, the method comprising the steps of: obtaining a first pulsed nmr log of the formation, the pulse sequence of the first nmr log comprising an initial 90° radio frequency pulse, followed by a series of 180° radio frequency pulses starting at a time period t_cp1 after the initial 90° pulse, and the series of pulses comprising magnetic pulses each separated by a time period 2t_cp1 ; obtaining a second pulsed nmr log of the formation, the pulse sequence of the second nmr log comprising an initial 90° radio frequency pulse, followed by a series of 180° pulses starting at a time period t_cp2 after the initial 90° pulse, and the series of 180° pulses comprising radio frequency pulses each separated by a time period 2t_cp2 wherein the time t_cp2 is a time that is different from t_cp1 by an amount of time sufficient to separate resultant peaks of transverse relaxation times attributable to the hydrocarbon phase or phases within the formation; and determining, from the first and the second nmr logs, the pore volume of a formation occupied by hydrocarbon phase or phases within the formation. In a preferred embodiment, pore volume occupied by hydrocarbon gas is determined, and t_cp2 is a time that is greater than that which results in the measured transverse relaxation time of the hydrocarbon gas being less than about 2×10^{-3 seconds and t_cp1 is a time that is less than that which results in the measured transverse relaxation time of the hydrocarbon gas being greater than about 4×10-3 seconds.}

Description

This application claims the benefit of U.S. Provisional Application No. 60/032,498, filed Dec. 18, 1996, the entire disclosure of which is hereby incorporated by reference

FIELD OF THE INVENTION

The invention relates to nuclear magnetic resonance logging of formations to estimate hydrocarbon content of the formation.

BACKGROUND TO THE INVENTION

In the exploitation of hydrocarbon reservoirs, use is made of well logging to determine the amount of recoverable hydrocarbons. Logging tools such as density, neutron, and resistivity logs, have been developed for measurement of reservoir properties such as porosity and water and hydrocarbon saturation of pore space. These tools are widely utilized in the oil industry. However, in order to accurately determine porosity using these tools, the lithology of the rock must be known. Many other rock and fluid properties, such as salinity, cementation factor, saturation exponent, and shaliness, must be known from sources such as resistivity logs to determine hydrocarbon saturation. In addition, methods are not known for estimating pore size or permeability in a continuous log, i.e. without taking fluid samples.

Nuclear Magnetic Resonance ("NMR") well logging tools capable of determining the liquid contents of pore volume within a reservoir, and a method to use these tools are described in, for example, U.S. Pat. Nos. 5,309,098, 5,291,137, 5,280,243, 5,212,447, 4,717,878, 4,717,877, 4,717,876, and 4,710,713. In particular, U.S. Pat. No. 5,291,137 discloses a Carr-Purcell-Meiboom-Gill ("CPMG") pulse sequence and echo response and a method to obtain free fluid porosity, total NMR porosity, bound fluid porosity, spin-spin or transverse relaxation time (T₂) (which is related to pore size distribution in sandstone), and continuous permeability logs. Recovery times between CPMG pulse trains is typically between 0.5 and 1.5 seconds. Because the spin lattice or longitudinal relaxation time T₁ of methane in typical reservoir conditions is greater than about three seconds, the total NMR porosity measured in this method could not include the volume occupied by hydrocarbon gas.

Recently, a new logging tool, the MRIL (TM NUMAR Corp. of Malvern, Pa.) has been introduced for determining the liquid-filled porosity in a lithology independent manner, i.e. the tool response does not require a lithology correction to determine porosity. The MRIL uses pulsed nuclear magnetic resonance of the mobile protons in the pore space. However, according to the tool's manufacturer, this tool is not capable of measuring the hydrocarbon gas content of the pore space.

This inability to detect gas has been a major disadvantage because the main or only hydrocarbon in many reservoirs is natural gas. The pore fluids in these reservoirs consist of only brine and natural gas. The inability to measure gas means that the NMR logging tool will not measure gas-filled porosity and must rely on comparison with other logging tools, such as density and neutron logs, to measure porosity.

It is therefore an object of the present invention to provide a method to determine the hydrocarbon content within a formation using a nuclear magnetic resonance ("NMR") log wherein formation properties, other than pore pressure and temperature, do not have to be known. It is another object of the present invention to provide a method to determine the hydrocarbon content that is not affected by the formation's clay content. It is a further object to provide such a method wherein the method is less susceptible to errors due to noise in the NMR log.

SUMMARY OF THE INVENTION

These and other objects are accomplished by a method to estimate the pore volume of a formation occupied by hydrocarbon phase or phases utilizing a NMR logging tool, the method comprising the steps of:

obtaining a first pulsed NMR log of the formation, the first NMR log utilizing a pulse sequence comprising an initial 90° radio frequency pulse, followed by a series of 180° radio frequency pulses starting at a time period t_cp1 after the initial pulse, and the series of 180° radio frequency pulses comprising 180° radio frequency pulses each separated by a time period 2t_cp1 wherein the time t_cp1 is a time which results in the measured transverse relaxation time of the hydrocarbon phase or phases being within a range of measured transverse relaxation times that are detectable with the NMR logging tool utilized;

obtaining a second pulsed NMR log of the formation, the pulse sequence of the second NMR log comprising an initial 90° radio frequency pulse, followed by a series of 180° radio frequency pulses starting at a time period t_cp2 after the initial 90° pulse, and the series of pulses comprising 180° radio frequency pulses each separated by an interecho time period 2t_cp2 wherein the time t_cp2 is a time that is different from t_cp1 by an amount of time sufficient to separate resultant peaks of transverse relaxation times attributable to hydrocarbon phase or phases within the formation between the two logs; and

determining from the first and the second NMR logs the pore volume of a formation occupied by hydrocarbon phase or phases.

In a preferred embodiment, formation porosity containing hydrocarbon gas is determined.

By adjusting the interecho time (2t_cp) of the NMR pulse sequence (a CPMG sequence) the transverse relaxation time T₂ of the hydrocarbon is significantly altered because of the strong dependence of T₂ on diffusional relaxation for gases. Liquid water, whether bound or free, is much less influenced by diffusion as a method of relaxation and therefore have NMR responses to CPMG sequences that are not significantly affected by altering the interecho time of the CPMG pulse sequence. A T₂ of less than about two milliseconds cannot be detected by current commercial NMR logging tools. Therefore logs using an interecho time sufficiently long that T₂ is less than two milliseconds will not indicate any peaks attributable to hydrocarbon gas. The difference between the NMR detected porosity of one such log using a long interecho time and another having an interecho time sufficiently short that T₂ is detected will therefore directly relate to gas-filled porosity. Similarly, a longer t_cp can be utilized, the longer t_cp resulting in both the liquid hydrocarbon phase, and the gas hydrocarbon phase being below deductibility, resulting in only the water-filled porosity being detected. Therefore either the liquid hydrocarbon, the hydrocarbon gas, or both can be determined by the differences between two NMR logs, the two NMR logs being run with differing t_cp s.

Current commercial NMR logging tools also have a lower limit of interecho times of about one millisecond. Use of an interecho time at about this lower limit results in hydrocarbon gas peaks of about forty milliseconds (for a magnetic field gradient of 17 gauss/cm). Utilization of shorter interecho times could further separate gas peaks from peaks attributable to liquids, and could be advantageous if a logging tool were to become available with such capabilities. The method of the present invention can be utilized to determine gas content of formations even with interecho times greater than one millisecond because, although hydrocarbon gas peaks may over-lap peaks from bound liquids, the differences between the peaks from the two logs would still indicate the amount of hydrocarbon gas present in the formation.

BRIEF DESCRIPTION OF THE FIGURES

FIG. 1 is a plot of T₁ of methane as a function of pressure for different temperatures.

FIG. 2 shows a typical CPMG echo sequence (with phase alternation of the 90° pulse) used for measuring transverse relaxation time, T₂.

FIG. 3 is a plot of hydrogen index for natural gas as a function of pressure for different temperatures.

FIG. 4 is a plot of T₂ decay curves for a gas bearing formation using interecho times of 1.2×10-3 and 4.8×10-3 seconds.

FIG. 5 is a plot of T₂ spectra obtained from an inversion of the T₂ decay curves of FIG. 4.

DETAILED DESCRIPTION OF THE INVENTION

The NMR log of the present invention is altered from that typically used in oil industry well logging by manipulation of the interecho time to change the response of hydrocarbon gas to the pulse sequence. A CPMG echo sequence with phase alteration of the 90° radio frequency pulses, such as is disclosed in U.S. Pat. No. 5,291,137 is utilized, along with a NMR logging tool such as the MRIL C available from NUMAR Corp. of Malvern, Pa. The NMR logging tool preferably is a tool that utilizes a magnetic gradient. The gradient of the MRIL C tool is about 17 gauss/cm. The gradient may be a pulsed or a fixed gradient. Such gradients are typically induced by magnets within the logging tool, but internal rock gradients may possibly be utilized. The MRIL tool is capable of sensing properties of a portion of a formation that is as much as four or five inches from the wellbore wall. A tool such as the MRIL C is therefore preferred in the practice of the present invention. This is preferred because rocks within less than two or three inches from the wellbore wall may be contaminated with drilling fluids and not at all representative of the formation in general. With the MRIL-C tool the two logs can be obtained on the same pass. MRIL-C tool has two different annuli located slightly apart from the other. These two different annuli can be activated with two different pulse sequences.

Referring now to FIG. 2, an exemplary CPMG pulse sequence with phase alternation of the 90° pulse is shown. This pulse sequence is used for measuring a distribution of transverse relaxation times, T₂. The sequence consists of a 90-degree radio frequency pulse (at the Larmor frequency), followed at time t_cp by a train of equally spaced 180-degree pulses. The spacing of the 180-degree pulses is 2t_cp. For example in the MRIL C this time can be as short as 1.2 ms. A spin echo, e, is obtained between each of the 180-degree pulses. The sequence is repeated after a wait time T_W with a 90° radio frequency pulse of opposite phase (relative to the preceding 90° radio frequency pulse). The subsequent train of negative echos, e', is subtracted from the previous train thus building up coherent signals and canceling instrument artifacts.

When there are multiple fluids in the pore space, and a range of pore sizes, the NMR signal as a function of time, A(t), represents a sum of exponential decays: ##EQU1## where a_i is a constant, T₂i is a constant representative of a relaxation time and n is an integer wherein n T₂i s are selected at equal logarithmic intervals. Typically, thirty five to fifty intervals, n, result in an acceptable fit to echo data. The time domain data can be inverted using a multiexponential inversion program yielding a histogram, or a plot of a_i as a function of T₂i. This inversion is discussed in, for example, U.S. Pat. No. 5,291,137. A multiexponential inversion of an echo train from a NMR response such as that displayed in FIG. 2 can therefore be expressed as a T₂ relaxation time distribution. The ordinate would be the signal amplitude associated with each T₂i time constant by fitting a_i for the given sets of T₂i s. NMR pore volumes can be determined by the integral of a_i with respect to T₂, and application of a constant determined from calibration, and multiplication with an effective hydrogen index.

Relaxation times of components that are bound to solids are generally significantly shorter than components that are not bound to solids. Surfaces in formations are generally either water or oil wet but are not gas wet. Thus, surface relaxation effects are negligible for gases.

Natural gas is composed predominantly of methane and light alkanes. Typically, over 75% by volume of dry gas is methane. Properties of the hydrocarbon gas within a formation can therefore be estimated with sufficient accuracy for the practice of the present invention by assuming a hypothetical hydrocarbon such as a C₁.1 H₄.2 hydrocarbon composition.

Longitudinal relaxation times, T₁, of gases such as methane are solely a function of temperature and pressure, and not of other properties of the formation. T₁ for methane is discussed in, for example, C. J. Gerritsma, et al., "Proton Spin Lattice Relaxation and Self Diffusion in Methanes-Paper 2", Physica, v. 5, 392 (1971). T₁ is considered to be proportional to the density and to vary with absolute temperature according to: ##EQU2## where: a and b are constants and T is the absolute temperature.

Referring now to FIG. 1, a plot of T₁ for a natural gas having a composition of C₁.1 H₄.2 is shown as a function of pressure for different temperatures. Equation 2 can be used to extrapolate T₁ data for methane and hydrocarbon gas compositions to other temperatures. Lines a through f represent T₁, in seconds, for temperatures of 100° F. through 350° F. in fifty degree increments, respectively.

As an example of typical reservoir conditions, methane may have a density of about 0.2 g/cc and a temperature of about 200° F., resulting in a T₁ of about four seconds. A wait time of six seconds will generally exceed T₁, and result in a NMR log that is useful in the practice of the present invention. Thus in order not to completely saturate the signal from methane the wait time (T_W) in the CPMG sequence (the time between the last 180° pulse and the next initial 90° pulse of the CPMG sequence) should be greater than four seconds, and preferably between about six and twelve seconds which is two to three times the T₁ of gas. T₁ of natural gas is between about 3 and 6 seconds for typical reservoir conditions.

In FIG. 2, the sequence is repeated after a wait time, T_W. If T_W is greater than three times T₁, then almost complete relaxation will occur. If T_W is not significantly greater than T₁, a correction factor, α, is applied to account for partial saturation. This correction is given by: ##EQU3##

Both logs of the present invention can be performed using a T_W significantly longer than T₁ so that substantial corrections for partial saturation will not have to be made. It is preferred that T_W of the two logging runs be the same but it is not necessary that they be the same. T_W for the log using a longer t_cp can be made using a relatively short T_W with respect to the gas T₁ (i.e., T_W less than T₁ of the hydrocarbon gas at formation conditions). The short T_W also suppresses the NMR response of the gas along with the suppression of the NMR response of the gas due to the changed t_cp.

NMR logging is normally restricted to measuring hydrogen (¹ H) because of its strong signal and high gyromagnetic ratio. Log results must therefore be corrected for hydrogen density to determine pore volumes. The hydrogen index, HI, is defined as the density of hydrogen atoms relative to liquid water at saturated conditions, and the HI is typically used to convert NMR porosities (raw results) to actual pore volumes. HI for hydrocarbon gases are known and available in, for example, Schlumberger Log Interpretation Principals/Applications, (1987), available from Schlumberger Educational Services, Houston, Tex., and in particular, FIGS. 5-17, p. 45, wherein density and HI of natural gas slightly heavier than methane (C₁.1 H₄.2) are shown as a function of pressure and temperature. Under typical reservoir conditions, the gas pressure will be between about 2000 and 10,000 psi and the temperature will be between about 100° F. and about 350° F., resulting in gas densities between about 0.1 and 0.3 g/cc and a HI between about 0.2 and about 0.6. Although this HI is less than one, it is still sufficiently large and renders the hydrocarbon gas measurable with the NMR pulse sequence of the present invention.

Referring now to FIG. 3, a plot is shown of HI for a natural gas having a composition of C₁.1 H₄.2 as a function of pressure for different temperatures. Lines g through l represent HI at temperatures of 100° F. through 350° F. in fifty degree increments respectively.

The "effective HI" is referred to herein as the product of α and HI.

The relaxation mechanisms that affect T₁ and T₂ in rocks are (1) molecular motion in fluids, (2) surface relaxivity at the pore wall, and (3) molecular diffusion in magnetic field gradients.

The first mechanism, due to local motions such as molecular tumbling, is called bulk relaxation. Longitudinal relaxation times and transverse relaxation times are equal when the predominant relaxation mechanism is bulk relaxation, i.e., T₁B ≈T₂B. Bulk relaxation is the predominant relaxation mechanism for non-wetting phases, i.e., T₁ =T₁B.

The second relaxation mechanism is surface relaxation at the pore wall, or the relaxation of ¹ H nuclei when they closely approach paramagnetic ions such as iron and manganese which reside on grain surfaces. This is the dominant mechanism for fluid molecules such as water that wet the rock surfaces. However, because gas is always non-wetting and never closely approaches the rock surface, this mechanism is negligible for gas. This provides a major advantage in the practice of the present invention in that only one relaxation time is measured for the gas, not a distribution of relaxation times that is a function of surface properties of the rock.

The third relaxation mechanism is the diffusion of molecules in magnetic field gradients. This relaxation mechanism affects only T₂ and not T₁. Therefore, when diffusion is a prominent relaxation mechanism, T₂ will be substantially less than T₁. Diffusion is the predominant relaxation mechanism for gas using a logging tool such as the MRIL-C.

Transverse relaxation time, T₂, in a magnetic field gradient is a function of bulk relaxation, surface relaxation, and diffusion according to: ##EQU4##

Using a CPMG sequence, the relaxation time due to diffusion (T₂D) is: ##EQU5## where γ is the ¹ H gyromagnetic ratio (26,741 radians/s-gauss), D is the diffusion coefficient, G is the field gradient created by the NMR tool, and t_CP is half of the interecho time. This is the predominate relaxation mechanism for the gas phase. From Equations 4 and 5, it can also be seen that t_cp could be increased to an extent that 1/T₂d is the dominate, or controlling relaxation mechanism.

The diffusion coefficient of a gas in a formation is limited on a upper limit by unrestricted diffusion, and at a lower limit by restricted diffusion.

The unrestricted diffusion coefficient, D_o of supercritical methane as a function of temperature and density can be found, for example, in Gerritsma et al., supra, and also in Dawson et al., AlChE Journal, Vol 16, No. 5, 1970. Under typical reservoir conditions, methane will have a density of about 0.2 g/cc, and a diffusion coefficient about 50 times that of water, or about 109×10-5 cm² /s compared to 2×10-5 cm² /s for water. Thus, using Equation (5), if unrestricted bulk diffusion occurred, with the MRIL C logging tool (G=17 gauss/cm, t_CP =0.6 ms), T₂D would equal 37.1 ms.

A ratio of the restricted diffusion coefficient, D', over the unrestricted diffusion coefficient, D_o approaches a limit of the inverse of the tortuosity as the fluids diffuse through many pores. Tortuosity is defined as the product of the formation resistivity factor, F, and the formation porosity, φ. The formation resistivity factor can be determined, for example, using an induction log, and the porosity can be determined from a neutron log.

In the practice of the present invention, it is not necessary to know the diffusion coefficient, D, with accuracy to determine the pore volume occupied with hydrocarbon gas because the hydrocarbon gas response is shifted by alteration of t_cp in the present invention. Pore volume filled by gas is identified by the difference between two NMR logs.

The range of transverse relaxation times within which a hydrocarbon gas peak is to be expected in an NMR log can therefore be anticipated using Equation 4 and Equation 5, because the hydrocarbon gas is not a surface-wetting phase, as: ##EQU6## Rearranging Equation 7 and recalling from above that T₁ =T₁B ≈T₂B : ##EQU7##

The interecho time t_cp can therefore be calculated that would result in the range within which T₂ for the hydrocarbon gas would be within a detectable range. For a MRIL C logging tool, with typical reservoir conditions, a t_cp of about 0.6 ms would result in a measured T₂ being about 40 ms. The NMR porosity indicated by this log would therefore include porosity filled with hydrocarbon gas. Another NMR log using a CPMG sequence having a different interecho time than the first NMR log preferably uses an t_cp that results in the T₂ due to the hydrocarbon gas being less about 2.5 ms. For a MRIL C logging tool, with typical reservoir conditions, a t_cp of about 2.4 ms would be sufficient to result in a T₂ below detectable limits of a tool such as the MRIL-C. The difference between the NMR porosities from two such logs, divided by the effective hydrogen index at formation conditions, would be the hydrocarbon gas content of the formation because the second log would not detect the hydrocarbon gas.

In order to remove T₂ bias, one should use only a subset of the echoes obtained under the short t_cp sequence which occur at the same times as the echoes obtained under the long t_cp sequence. Thus if t_cp2 is four times t_cp1, one would use only every fourth echo in the t_cp1 sequence.

In the practice of the present invention, it is not necessary that a multiexponential inversion be performed on echo data obtained from the NMR logging tool. Rather than the results of the multiexponential inversion, subtracting integrals of the time domain echo responses yields a result that is proportional to the gas-filled porosity, φ_g, (when an interecho time is used that is long enough to suppress T₂ of the gas peak below detection in one of the two logs, 2t_cp2, and an interecho time short enough to result in a measurable T₂ of the hydrocarbon gas peak, 2t_cp1). Utilizing the difference between the integrals of the time domain echo responses is preferable because noise is more directly canceled. This can be seen starting with equations for measured decay curves as functions of time:

A₁ (t)=S₁ (t)+N₁ (t) (9)

A₂ (t)=S₂ (t)+N₂ (t) (10)

where: A₁ (t) is the measured decay curve obtained using t_cp1 ;

A₂ (t) is the measured decay curve obtained using t_cp2 ;

S₁ (t) is the signal of the decay curve without noise for the log using t_cp1 ;

S₂ (t) is the signal of the decay curve without noise for the log using t_cp2 ;

N₁ (t) is noise in the measured decay curve using t_cp1 ; and

N₂ (t) is noise in the measured decay curve using t_cp2.

Noise is assumed to be random thermal noise which has the same expectation value on both passes. Therefore:

∫N₁ (t)dt-∫N₂ (t)dt=0 (11)

Integrating the two measured decay curves A₁ (t) and A₂ (t) from a time t₀, which is much less than T₂ of the hydrocarbon gas, T₂g, with t_cp1, and greater than T₂ of the hydrocarbon gas with t_cp2, to T₀ which is much greater than T₂g at t_cp1 results in the cancellation of the integrals of N₁ (t) and N₂ (t) according to: ##EQU8## Because the only change between the two logging scans is the interecho time 2t_cp which was altered to make hydrocarbon gas undetectable in one of the two logging scans, the difference between the two measured decay curves is the decay curve attributable to the hydrocarbon gas decay. This hydrocarbon gas decay curve, g(t), is one of the decay curves of the summation of Equation 1, which can be expressed as: ##EQU9##

Integration of this function from t₀ to T₀ results in: ##EQU10## Because to is much less than T₂g and T₀ is much greater than T₂g : ##EQU11##

Therefore, Equations 14 and 12 can be solved for φ_g : ##EQU12##

Another alternative is to extrapolate an echo decay curve to time (or number of echo) equal to zero. The signal amplitude at zero time is also proportional to NMR porosity. These alternatives eliminate the need to do the multiexponential inversion.

The TABLE below lists typical T₁, T₂, HI, D_o and D_o T₁ for brine, oil and natural gas. The values of T₂ are for the MRIL-C tool with a gradient of about 17 gauss/cm and t_cp of 0.6 ms. The TABLE shows that the T₁ of oil and natural gas may overlap despite a large T₂ contrast. One can also see from the TABLE that brine and oil may have overlapping D_o or T₂, but distinctly separated T₁. Because D_o T₁ of gas is an order of magnitude greater than oil, and two orders of magnitude greater than brine, one can see from Equation 8 that the already large DT₁ contrast of gas can be enhanced by increasing the interecho time 2t_cp in order to allow the separation of two fluids (such as oil and hydrocarbon gas) that overlap in T₁.

TABLE
______________________________________
T1   T2           D0 × 10-5
D0 T1
ms        ms        HI      cm2 /s
cm2
______________________________________
brine
1 to 500 0.67 to 200
1     7.7    0.0077 to 4
oil  5000     460       1     7.9     40
gas  4400      40       0.38  100    440
______________________________________

From the TABLE and Equation 8, it can be seen that the present invention can be applied to determine porosities filled with liquid hydrocarbons, or hydrocarbon gas plus liquid hydrocarbons by increasing t_cp to a time that causes T₂ of both the hydrocarbon gas and the liquid hydrocarbon to shift. The significant difference between D_o T₁ of brine and oil indicates that a range of t_cp s could be determined wherein the T₂ peak of the liquid hydrocarbons has shifted significantly whereas the T₂ peak of brine is still relatively unaffected compared to a NMR log using a different t_cp. Thus the method of the present invention can be used to determine the porosity of a formation filled with gas, liquid hydrocarbon, or gas plus liquid hydrocarbon by simply integrating time domain responses of NMR logs using different t_cp s.

When compared to prior art method of determine hydrocarbon liquid content of a formation, i.e., performing a multiexponential inversion of time domain responses and distinguishing based on the location of T₂ peaks, the present method more directly cancels noise. Further, considerably less computing capacity or data storage is required because of elimination of the multiexponential inversion.

The MRIL C tool has a depth-of-investigation of 16 inches. In an eight inch diameter borehole, the formation can therefore be examined up to four inches of the bore hole wall Oil-based muds have low invasion to this depth and are therefore the preferred drilling mud for practice of this invention. With low invasion oil-based muds the gas saturation observed by the MRIL C logging tool will be substantially unflushed. In particular, ESCAID 110 oil-based drilling mud with 80% ESCAID 110 and 20% CaCl2-saturated water to provide very little invasion in Gulf of Mexico sands and is therefore a preferred system.

EXAMPLE

A numerical simulation was performed of two logs of a formation with a MRIL C tool, with a T_W of eight seconds, and t_cp 's of 0.6 and 2.4 ms. FIG. 4 is a plot of echo vs. echo amplitude for both logs, with line "a" representing the log with t_cp of 0.6 ms and line "b" representing the log with t_cp of 2.4 ms. The formation was a gas reservoir at irreducible conditions with capillary bound water-filled porosity of 10 p.u. and gas-filled porosity of 20 p.u. The NMR properties for the gas phase are those in the TABLE. Only every fourth echo from the log using a t_cp of 0.6 ms was used to eliminate T₂ bias. Also, the first two echoes from the resulting CPMG sequences were discarded.

FIG. 5 is a plot of NMR porosity, in percent, vs. T₂ calculated from the decay curves of FIG. 4, with lines a and b representing the porosity measured using 0.6 and 2.4 ms interecho times respectively. It can be seen from FIG. 5 that a distinct peak from the gas phase centered at about 40 ms is prevalent in the log using the t_cp of 0.6 ms whereas this peak is not present in the log using the t_cp of 2.4 ms. From FIG. 5 it can be seen that the longer t_cp has shifted the gas spectrum below deductibility, leaving only the brine signal, centered at 12 ms, in the spectrum. The difference between the integrals of the area under these curves therefore represents the NMR porosity attributable to hydrocarbon gas.

Claims

1. A method to estimate the formation pore volume occupied by hydrocarbon gasp the method comprising the steps of:

obtaining a first pulsed nmr log of the formation, the first nmr log utilizing a pulse sequence comprising an initial 90° radio frequency pulse, followed by a series of 180° radio frequency pulses starting at a time period t_cp2 after the initial 90° pulse, and the series of 180° pulses comprising radio frequency pulses each separated by a time 2t_cp2 wherein the time t_cp2 is a time that is greater than that which results in the measured transverse relaxation time of the hydrocarbon gas being less than about 2×10^{-3 seconds;}

obtaining a second pulsed nmr log of the formation, the second nmr log utilizing a pulse sequence comprising an initial 90° radio frequency pulse, followed by a series of 180° pulses starting at a time period t_cp1 after the initial 90° pulse, and the series of 180° pulses comprising radio frequency pulses each separated by a time 2t_cp1 wherein the time t_cp1 is a time that is less than that which results in the measured transverse relaxation time of the hydrocarbon gas being greater than about 4×10^{-3 seconds;}

determining from the first and the second nmr logs the pore volume of a formation occupied by hydrocarbon gas.

2. The method of claim 1 wherein the pore volume of the formation occupied by hydrocarbon gas is determined by:

determining, from the first and the second nmr logs, a distribution of transverse relaxation times attributable to hydrocarbon gas within the formation; and

determining from the distribution of transverse relaxation times attributable to hydrocarbon gas within the formation the pore volume of a formation occupied by hydrocarbon gas within the formation.

3. The method of claim 1 wherein t_cp1 is less than that which results in the measured transverse relaxation time of the hydrocarbon gas being greater than about 8×10^{-3 seconds.}

4. The method of claim 1 wherein t_cp2 is equal to or greater than about 2.4 ms.

5. The method of claim 1 wherein t_cp1 is equal to or less than about 0.6 ms.

6. The method of claim 1 wherein the two logs are obtained at the same time by a tool utilizing different pulse sequences in two annuli located slightly apart from each other.

7. The method of claim 1 wherein both the first and the second log are obtained using pulse sequences wherein t_W is about six seconds or more.

8. The method of claim 2 wherein t_cp2 is equal to or greater than about 2.4 ms.

9. The method of claim 8 wherein t_cp1 is equal to or less than about 0.6 ms.

10. The method of claim 9 wherein the two logs are obtained at the same time by a tool utilizing different pulse sequences in two annuli located slightly apart from each other.

11. The method of claim 10 wherein both the first and the second log are obtained using pulse sequences wherein t_W is about six seconds or more.

12. A method to estimate a formation pore volume occupied by hydrocarbon gas utilizing a nmr logging tool, the method comprising the steps of:

obtaining a first pulsed nmr log of the formation, the first nmr log utilizing a pulse sequence comprising an initial 90° radio frequency pulse, followed by a series of 180° radio frequency pulses starting at a time period t_cp1 after the initial 90° pulse, and the series of 180° pulses comprising radio frequency pulses each separated by a time 2t_cp1 wherein the time t_cp1 is a time which results in the measured transverse relaxation time of the hydrocarbon gas being within a range of measured transverse relaxation times that are detectable with the nmr logging tool utilized;

obtaining a second pulsed nmr log of the formation, the second nmr log utilizing a pulse sequence comprising an initial 90° radio frequency pulse, followed by a series of 180° radio frequency pulses starting at a time period t_cp2 after the initial pulse, and the series of 180° pulses comprising radio frequency pulses each separated by a time period 2t_cp2 wherein the time t_cp2 is a time that is different from t_cp1 by an amount of time sufficient to separate resultant peaks of transverse relaxation times attributable to hydrocarbon gas within the formation; and

determining from the first and the second nmr logs the formation pore volume occupied by hydrocarbon gas.

13. The method of claim 12 wherein the formation pore volume occupied by hydrocarbon gas is determined by:

determining, from the first and the second nmr logs, a distribution of transverse relaxation times attributable to hydrocarbon gas within the formation; and

determining from the distribution of transverse relaxation times attributable to hydrocarbon gas within the formation the pore volume of a formation occupied by hydrocarbon gas within the formation.

14. The method of claim 12 wherein t_cp1 is less than that which results in the measured transverse relaxation time of the hydrocarbon gas being greater than about 8×10^{-3 seconds.}

15. The method of claim 12 wherein t_cp2 is equal to or greater than about 2.4 ms.

16. The method of claim 12 wherein t_cp1 is equal to or less than about 0.6 ms.

17. The method of claim 12 wherein the two logs are obtained at the same time by a tool utilizing different pulse sequences in two annuli located slightly apart from each other.

18. The method of claim 12 wherein both the first and the second log are obtained using pulse sequences wherein t_W is about six seconds or more.

19. The method of claim 12 wherein t_cp2 of the second log is sufficiently long that the measured transverse relaxation time of the hydrocarbon gas is below detectable limits of the nmr logging tool.

20. The method of claim 12 wherein the second log is performed using a t_W that is less than t₁ of the hydrocarbon gas at formation conditions.

21. A method to estimate a formation pore volume occupied by a hydrocarbon phase or phases utilizing a nmr logging tool, the method comprising the steps of:

obtaining a first pulsed nmr log of the formation, the first nmr log utilizing a pulse sequence comprising an initial 90° radio frequency pulse, followed by a series of 180° radio frequency pulses starting at a time period t_cp1 after the initial 90° pulse, and the series of 180° pulses comprising radio frequency pulses each separated by a time 2t_cp1 wherein the time t_cp1 is a time which results in the measured transverse relaxation time of the hydrocarbon phase or phases being within a range of measured transverse relaxation times that are detectable with the nmr logging tool utilized;

obtaining a second pulsed nmr log of the formation, the second nmr log utilizing a pulse sequence comprising an initial 90° radio frequency pulse, followed by a series of 180° radio frequency pulses starting at a time period t_cp2 after the initial pulse, and the series of 180° pulses comprising radio frequency pulses each separated by a time period 2t_cp2 wherein the time t_cp2 is a time that is different from t_cp1 by an amount of time sufficient to separate resultant peaks of transverse relaxation times attributable to the hydrocarbon phase or phases within the formation; and

determining from the first and the second nmr logs the formation pore volume occupied by the hydrocarbon phase or phases.

22. The method of claim 21 wherein the formation pore volume occupied by the hydrocarbon phase or phases is determined by:

determining, from the first and the second nmr logs, a distribution of transverse relaxation times attributable to hydrocarbon phase or phases within the formation; and

determining from the distribution of transverse relaxation times attributable to hydrocarbon phase or phases within the formation the pore volume of a formation occupied by hydrocarbon gas within the formation.

23. The method of claim 21 wherein t_cp1 is less than that which results in the measured transverse relaxation time of the hydrocarbon gas being greater than about 8×10^{-3 seconds.}

24. The method of claim 21 wherein t_cp2 is equal to or greater than about 2.4 ms.

25. The method of claim 21 wherein t_cp1 is equal to or less than about 0.6 ms.

26. The method of claim 21 wherein the two logs are obtained at the same time by a tool utilizing different pulse sequences in two annuli located slightly apart from each other.

27. The method of claim 21 wherein both the first and the second log are obtained using pulse sequences wherein t_W is about six seconds or more.

28. The method of claim 21 wherein t_cp2 of the second log is sufficiently long that the measured transverse relaxation time of the hydrocarbon phase or phases is below detectable limits of the nmr logging tool.

29. The method of claim 28 wherein the formation pore volume occupied by the hydrocarbon phase or phases is determined as a function of the difference between the integrals of the time domain responses of the two nmr logs.